Spatial Risk Analysis for the Introduction and Circulation of Six

Esser et al. Parasites Vectors (2020) 13:464 https://doi.org/10.1186/s13071-020-04339-0 Parasites & Vectors

RESEARCH Open Access Spatial risk analysis for the introduction and circulation of six arboviruses in the Netherlands Helen Joan Esser1,2,3*† , Yorick Liefting1†, Adolfo Ibáñez‑Justicia4, Henk van der Jeugd5, Chris A. M. van Turnhout6,7, Arjan Stroo4, Chantal B. E. M. Reusken3,8, Marion P. G. Koopmans8 and Willem Fred de Boer1

Abstract Background: Arboviruses are a growing public health concern in Europe, with both endemic and exotic arboviruses expected to spread further into novel areas in the next decades. Predicting where future outbreaks will occur is a major challenge, particularly for regions where these arboviruses are not endemic. Spatial modelling of ecological risk factors for arbovirus circulation can help identify areas of potential emergence. Moreover, combining hazard maps of diferent arboviruses may facilitate a cost-efcient, targeted multiplex-surveillance strategy in areas where virus trans‑ mission is most likely. Here, we developed predictive hazard maps for the introduction and/or establishment of six arboviruses that were previously prioritized for the Netherlands: West Nile virus, Japanese encephalitis virus, Rift Valley fever virus, tick-borne encephalitis virus, louping-ill virus and Crimean-Congo haemorrhagic fever virus. Methods: Our spatial model included ecological risk factors that were identifed as relevant for these arboviruses by an earlier systematic review, including abiotic conditions, vector abundance, and host availability. We used geo‑ graphic information system (GIS)-based tools and geostatistical analyses to model spatially continuous datasets on these risk factors to identify regions in the Netherlands with suitable ecological conditions for arbovirus introduction and establishment. Results: The resulting hazard maps show that there is spatial clustering of areas with either a relatively low or rela‑ tively high environmental suitability for arbovirus circulation. Moreover, there was some overlap in high-hazard areas for virus introduction and/or establishment, particularly in the southern part of the country. Conclusions: The similarities in environmental suitability for some of the arboviruses provide opportunities for targeted sampling of vectors and/or sentinel hosts in these potential hotspots of emergence, thereby increasing the efcient use of limited resources for surveillance. Keywords: Risk mapping, Geographic Information System, West Nile virus, Japanese encephalitis virus, Rift Valley fever virus, Tick-borne encephalitis virus, Louping-ill virus, Crimean-Congo haemorrhagic fever virus, Vector-borne diseases

Background Arthropod-borne viruses, or arboviruses, are an increas- *Correspondence: [email protected] ing public health concern in Europe [1]. Diseases caused †Helen Joan Esser and Yorick Liefting contributed equally to this work 1 Wildlife Ecology & Conservation Group, Wageningen University & by endemic arboviruses such as West Nile virus (WNV), Research, Wageningen, The Netherlands Crimean-Congo haemorrhagic fever virus (CCHFV), and Full list of author information is available at the end of the article tick-borne encephalitis virus (TBEV) are all increasing

© The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativeco​ ​ mmons.org/licen​ ses/by/4.0/​ . The Creative Commons Public Domain Dedication waiver (http://creativeco​ mmons​ .org/publi​ cdoma​ in/​ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Esser et al. Parasites Vectors (2020) 13:464 Page 2 of 20

in incidence and distribution [2–4]. Moreover, the wide- vectors, reservoir hosts, and abiotic conditions that need spread occurrence and periodically high local abun- to converge for arbovirus replication and transmission dance of competent vectors and reservoir hosts increase to occur [22]. Spatial modelling of these ecological risk the probability that exotic arboviruses, such as Japanese factors can be used to map the environmental suitability encephalitis virus (JEV) or Rift Valley fever virus (RVFV), (‘hazard’) for local circulation of arboviruses in regions become established [5, 6]. Projected changes in land-use beyond their current distribution [23–27]. and climate, socio-economic development, and virus In this study, we developed predictive hazard maps evolution may all contribute to larger and more frequent for the introduction and/or establishment of the six ear- outbreaks in endemic regions, and promote geographic lier prioritized arboviruses TBEV, LIV, CCHFV, JEV, expansion of arboviruses, including those that were tra- WNV and RVFV in the Netherlands. Our spatial analy- ditionally confned to tropical regions, into novel areas sis included ecological risk factors that were identifed within Europe [7–10]. as relevant for these arboviruses by an earlier systematic Te above-mentioned mosquito-borne viruses (JEV, review, including abiotic conditions, vector abundance, WNV and RVFV) and tick-borne viruses (CCHFV, TBEV and host availability [28]. We used geographic informa- and the closely related louping-ill virus LIV) have been tion system (GIS)-based tools and geostatistical analyses marked as top priority arboviruses for the Netherlands to model spatially continuous data on these risk factors based on epidemiological criteria, and their economic to identify regions with suitable ecological conditions and societal impact [11]. Teir potential emergence for endemic circulation. It is in these potential hotspots in the Netherlands may be facilitated by the country’s where surveillance eforts should be focused to enable unique combination of (i) high densities of livestock, early risk management. Our specifc goal was to contrib- which function as reservoir hosts for JEV, RVFV, CCHFV ute to an integrated, multiplex surveillance strategy in and LIV, (ii) great global connectivity in trade and travel which hazardous areas are monitored for the presence of through large airports and seaports, which increases the multiple arboviruses simultaneously. risk for arbovirus introduction, and (iii) priority pol- icy to improve ecological conditions attracting wildlife Methods (both resident and migratory) via habitat conservation We considered Additional fle 1: Table S1 from the sys- and the establishment of wildlife corridors. Examples of tematic literature review by Esser et al. [22] to identify these conservation initiatives are the National Ecologi- ecological risk factors associated with sustained circula- cal Network (NEN), the Natura 2000 network, and the tion and spread of the six previously prioritized arbovi- Pan-European Ecological Network (PEEN), which are all ruses (JEV, WNV, RVFV, CCHFV, TBEV and LIV). Tese aimed at higher wildlife mobility and larger distribution factors included abiotic conditions (i.e. temperature, ranges, and hence may facilitate the spread of arboviruses humidity and precipitation), vegetation cover, and the with wildlife reservoirs. Moreover, increasingly warmer abundance of vectors and (reservoir) hosts such as migra- summers and milder winters improve the climatic suit- tory birds, livestock and deer. We then used nationwide ability of the Netherlands for the establishment of arbo- continuous data on these ecological factors to construct viruses and their vectors [12, 13]. Te recent outbreaks hazard maps for the potential introduction and/or estab- of Usutu virus in blackbirds and the frst autochthonous lishment of these arboviruses in the Netherlands. Factors human cases of TBE in this country, underline the con- that were not relevant for this country (e.g. elevation, tinuous threat of arbovirus emergence [14, 15]. presence of rice felds), or for which detailed information While history has shown that preventing the intro- was not publicly available (e.g. point-to-point interna- duction and spread of arboviruses and their vectors into tional livestock transport), were excluded from analysis. novel areas may be near impossible [16–19], potential Te available data allowed us to construct 9 hazard maps, outbreaks can be prevented or their efects mitigated by 3 for the introduction of CCHFV, WNV, and JEV, and 6 targeted early warning surveillance, preparedness plan- for the establishment of each arbovirus (see below). ning, and control eforts (e.g. monitoring of vectors and/ or sentinel hosts, vaccination, education campaigns and Risk factors for introduction biological control) in hazardous areas [20, 21]. However, Long-distance dispersal of infected vectors and/or wild- predicting where arbovirus emergence is most likely to life is an important mechanism for the introduction and occur and hence, selecting locations where surveillance spread of arboviruses into novel areas [28]. We consid- should be most efective, is a major challenge. A spatial ered the relative abundance of bird species that migrate analysis of ecological risk factors for arbovirus circulation from Africa and/or the Mediterranean area to the Neth- can help identify such areas. Arboviruses are associated erlands in spring as principal risk factor for the introduc- with the presence, abundance and interactions of specifc tion of WNV and CCHFV (see Additional fle 1: Table S1, Esser et al. Parasites Vectors (2020) 13:464 Page 3 of 20

for a list of included bird species). As these birds have role in RVFV epidemiology [43, 44], whereas the role of overwintering and/or stopover sites in endemic regions, small mammals and ruminant wildlife, such as deer, in they may potentially be viraemic with WNV or carry virus maintenance remains unclear [45, 46]. Tese ani- Hyalomma ticks infected with CCHFV upon arrival in mals were therefore not included as risk factors for the the Netherlands [29, 30]. Indeed, even though there are establishment of RVFV. For JEV, Ardeid birds and pigs no established populations of Hyalomma ticks in the are the principal reservoir hosts [33]. Hence, we included Netherlands, immature ticks are incidentally imported by the local abundance of Ardeid bird species that occur migratory birds [31, 32]. Migratory birds may also spread in the Netherlands (see Additional fle 1: Table S3) and JEV via overlapping migratory fyways with birds migrat- the abundance of pigs as risk factors for JEV establish- ing from JEV-endemic regions in southeast Asia [33]. ment. West Nile virus is a multi-host pathogen that is We therefore included the relative abundance of these maintained in a bird-mosquito transmission cycle [47]. bird species as risk factor for the introduction of JEV (see Although the reservoir competence of many European Additional fle 1: Table S2). bird species for WNV remains largely unknown, past Long-distance dispersal of infected livestock is argu- outbreaks have often occurred near wetland areas, where ably the most important mechanism for the introduction large numbers of wetland birds and ornithophilic mos- of RVFV and LIV into novel areas [28, 34]. In addition, quitoes concentrate [48, 49]. In addition, experimental livestock imported from CCHFV-endemic regions may studies have shown that the European carrion crow (Cor- also carry infected Hyalomma ticks [35]. However, data vus corone), the jackdaw (Coloeus monedula), the mag- on point-to-point international transport of livestock is pie (Pica pica), the rock pigeon (Columba livia) and the commercially and socially sensitive and was therefore not house sparrow (Passer domesticus) are all highly suscepti- made available by competent bodies for this study, which ble to WNV infection and are competent reservoir hosts precluded the construction of introduction risk maps for [50–54]. We therefore included the relative abundance of RVFV and LIV. Host movement also plays an important wetland bird species and that of crow, jackdaw, magpie, role in the spread of TBEV throughout Europe [36, 37], rock pigeon and house sparrow as risk factors for WNV but as this virus emerged in the Netherlands while this establishment (see Additional fle 1: Table S4). study was ongoing, we limited our analysis to estimating Temperature, humidity, and precipitation have all been the establishment of TBEV. associated with outbreaks of the three mosquito-borne viruses that we consider here [22]. However, we included Risk factors for establishment only temperature as abiotic risk factor because of its For the mosquito-borne viruses, we included the fol- direct impact on vector competence, biting rates, and the lowing risk factors for establishment: (i) abundance extrinsic incubation period of the virus [55, 56]. Other of competent vectors; (ii) abundance of competent abiotic factors, such as precipitation or humidity, are reservoir hosts; and (iii) suitable climatic conditions indirectly related to virus circulation via their impact on (Table 1). Competent mosquito species that have been mosquito abundance [57, 58], which we have modelled prioritized as a veterinary and public health concern separately. Specifcally, we included the average daily for the Netherlands, based on their local occurrence temperature during spring and summer (April to Octo- and vector status, include Aedes vexans for RVFV ber) as risk factor for WNV and RVFV establishment, and Culex pipiens for WNV and RVFV [38–40]. More with higher temperatures corresponding with higher risk recently, European Cx. pipiens mosquitoes were also [56, 59]. For JEV, we included the number of days when shown to be competent vectors for JEV [5]. We esti- the average daily temperature was at least 25 °C, a tem- mated the abundance of these two mosquito species perature limit above which JEV outbreaks occur [57]. across the Netherlands using random forest models For the tick-borne viruses, we included the following following the methods described in Ibañez-Justicia & risk factors for establishment: (i) suitable tick habitat; (ii) Cianci [41] and using the data reported in Ibañez-Jus- presence of key host species for adult ticks and/or virus ticia et al. [42]. Detailed information on these methods transmission; and (iii) abiotic conditions that facilitate and the ecological variables used can be found in Addi- virus transmission and/or tick development. Ticks are tional fle 2: Text S1. very sensitive to abiotic conditions and their survival is Diferent reservoir hosts are involved in the transmis- directly related to vegetation cover and leaf litter, which sion cycle of each of the three mosquito-borne viruses. protects them from desiccation or freezing [60]. Te Domestic ruminants are the most important reservoir principal vector of TBEV and LIV in Europe, Ixodes rici- host for RVFV [34]. We therefore included the abun- nus, occurs in a wide range of habitats, but it is typically dance of cattle, sheep, and goats as risk factor for RVFV found in woodlands and forests with thick undergrowth establishment. Pigs and horses do not play a signifcant [61, 62]. Indeed, TBE incidence is positively correlated Esser et al. Parasites Vectors (2020) 13:464 Page 4 of 20 - borne viruses (TBEV, Vector abundance and Aedes Culex abundance Culex abundance Culex Suitable habitat for I. ricinus ticks Suitable habitat for Suitable habitat for I. ricinus ticks Suitable habitat for ticks H. marginatum Suitable habitat for ranging livestock (i.e. (i.e. livestock - ranging borne viruses (i.e. RVFV, JEV and WNV) JEVand RVFV, - borne(i.e. viruses tick and three goat, and cattle) pigs magpie, pigeon, and house sparrow magpie, cattle, sheep, goat, horse) sheep, cattle, goat) Host Abundance of ruminant livestock (i.e. sheep, sheep, (i.e. Abundance of ruminant livestock Abundance of ardeid bird species and domestic bird Abundance of ardeid Abundance of wetland birds, crow, jackdaw, jackdaw, crow, birds, Abundance of wetland Presence of deer and free Presence Abundance of sheep Abundance of livestock (i.e. cattle, horse, sheep, sheep, horse, cattle, (i.e. Abundance of livestock Establishment risk factors Abiotic Positive efect of TG from April October to from TG of efect Positive Number of days with TX ≥ 25 °C with Number of days Positive efect of TG from April October to from TG of efect Positive Slope of TG decrease from August to October to August from decrease TG Slope of May to March from increase TG Slope of April October to of UG from efect Positive Slope of TG decrease from August to October to August from decrease TG Slope of May to March from increase TG Slope of April October to of UG from efect Positive Negative efect of RH efect Negative April October to from TG of efect Positive endemic - fyways with conspecifcs from JEVfyways with conspecifcs from areas in Asia areas Africa and/or the Mediterranean area to the Africa to and/or the Mediterranean area Netherlands in spring Africa and/or the Mediterranean area to the Africa to and/or the Mediterranean area Netherlands in spring Ecological risk factors associated with the spread and sustained circulation of three mosquito three of circulation sustained and risk the spread with factors associated Ecological Introduction risk factors na Abundance of birds with overlapping migratory with overlapping Abundance of birds Abundance of bird species that migrate from from species that migrate Abundance of bird na na Abundance of bird species that migrate from from species that migrate Abundance of bird - borne viruses - borne viruses RVFV JEV WNV TBEV LIV CCHFV 1 Table [ 22 ] of Esser et al. review factors based upon the systematic listed was of the below Choice TX, 24 h maximum temperature temperature; 24 h average TG, humidity; relative RH, 24 h sum of precipitation; : UG, 24 h average Abbreviations LIV and CCHFV) that were included for the development of hazard maps for arbovirus and establishment introduction LIV and CCHFV) maps for of hazard the development included for that were Mosquito Tick Esser et al. Parasites Vectors (2020) 13:464 Page 5 of 20

with the proportion of broad-leafed, mixed, and conif- spring as a risk factor for establishment of TBEV and LIV. erous forest stands, which also provide habitat for small Autumnal cooling was calculated as the slope of the aver- rodents that function as amplifying hosts for TBEV [63– age daily temperature decrease from August 1st to Octo- 66]. In contrast, the most prominent European vector of ber 31st. Spring warming was calculated as the slope of CCHF V, Hyalomma marginatum, prefers open country the average daily temperature increase from March 1st habitat [62], with clinical cases of CCHF being positively to May 31st. As moist conditions are a controlling fac- correlated with the proportion of shrub or grassland tor for the survival of I. ricinus, we also included a posi- cover and with habitat fragmentation in agricultural tive relationship with relative humidity [73]. In contrast, areas [8, 26, 67]. We therefore included diferent land- H. marginatum ticks are adapted to the warm climatic use types for I. ricinus and H. marginatum in our analysis conditions of northern Africa and southern Europe [74]. and scored these on a scale of 1 to 3, with higher values Various modelling studies showed a northward shift in corresponding to more suitable habitat (see Additional climate suitability of this species with increasing temper- fle 1: Table S5). atures and decreasing rainfall as predicted under future Large herbivores are fnal hosts of adult I. ricinus and climate change scenarios [75–77]. We therefore included H. marginatum ticks, and are also directly involved a negative relationship with rainfall and a positive rela- as reservoir host in the transmission cycle of LIV and tionship with temperature during summer months (April CCHFV [35, 36, 68]. In the Netherlands, deer presence to October) as risk factors for CCHFV establishment. rather than abundance best explains I. ricinus density [69]. In areas where deer and other wild herbivores are Raw source data absent, free-ranging livestock that are used for nature We used climatic data from the Royal Netherlands Mete- management may instead maintain tick populations by orological Institute (KNMI). Daily meteorological data of feeding adult ticks [70]. We therefore used the presence 38 stations from 1 January 2010 until 31 December 2015 of deer (i.e. roe deer Capreolus capreolus, fallow deer were used to interpolate (Spline function) daily maps Dama dama, red deer Cervus elaphus) and free-ranging with full coverage of the Netherlands. We followed the livestock (i.e. cattle, sheep, goat and horse) as a risk fac- KNMI protocol for interpolating daily meteorological tor for high densities of I. ricinus and hence local circula- data [78]. We included the following factors: 24 h average tion of TBEV. Louping ill virus most commonly occurs in temperature (TG; 0.1 °C); 24 h maximum temperature upland habitats of the British Isles, where the red grouse (TX; 0.1 °C), 24 h sum of precipitation (RH; 0.1 mm); and (Lagopus lagopus scotica) is a competent transmission 24 h average relative humidity (UG; %). To identify suit- host, and the mountain hare (Lepus timidus) supports able tick habitat, we used the LGN7 dataset [79] for land- all three life stages of I. ricinus as well as non-viraemic use in the Netherlands, which diferentiates between 39 transmission of LIV via co-feeding ticks [68]. Neither land-use types (Additional fle 1: Table S5). Livestock red grouse nor mountain hares are present in the Neth- abundance data was obtained from the 2015 livestock erlands. Teir absence, however, does not preclude local survey database (Landbouwtelling 2015; poll date 1 April circulation of LIV in this country; sheep are highly com- 2015) as provided by the Netherlands Enterprise Agency petent reservoir hosts and are capable of maintaining an (RVO). Presence of free-ranging livestock in nature enzootic cycle with I. ricinus ticks, even in the absence of reserves was provided by Wageningen Environmental other key hosts such as deer [68]. We therefore included Research. Data on the presence of roe deer and hares sheep abundance as a risk factor for the establishment of was obtained from the Dutch National Database Flora LIV. Livestock and hares are also principal host species and Fauna [80]. Data on the abundance of birds during for adult H. marginatum ticks and act as amplifying hosts the breeding season (spring) were obtained from the Bird for CCHFV [35]. Since the European hare (Lepus euro- Atlas of the Netherlands, based on nationwide feldwork paeus) is present throughout the Netherlands, but local in 2013–2015 (www.vogel​atlas​.nl). Te abundance of abundance data were not available, we only included the rare bird species was estimated per 5 × 5 km grid square abundance of livestock (i.e. cattle, goat, sheep and horse) on a semi-quantitative ordinal scale (classes: 1–3; 4–10; as a risk factor for CCHFV establishment. 11–25; 26–100; 101–500; 501–1000 breeding pairs). Rapid autumnal cooling followed by rapid spring Te abundance of common bird species was quantifed warming are considered to be key climatic conditions by using geostatistical modelling, based on bird counts for the transmission of TBEV and LIV as it enables syn- during standardized timed visits in eight systematically chronous activity of, and hence co-feeding transmission selected 1 × 1 km grid squares per 5 × 5 km square, and between, infected I. ricinus nymphs and uninfected lar- a set of environmental variables. For details of feld work vae [71, 72]. We therefore included the rate with which methods and modelling techniques, see Sovon Vogelond- temperatures decreased in autumn and increased in erzoek Nederland [81]. As the relative importance of Esser et al. Parasites Vectors (2020) 13:464 Page 6 of 20

each bird species for virus transmission remains unclear, layers were then averaged and normalized again to con- we weighted each bird species equally in our analyses. For struct the establishment map (see Additional fle 4: Fig- rare bird species, we took the geometric mean per abun- ures S20–S25, for a schematic representation of the dance class, and then normalized the abundance of each procedure). species between 0 and 100. For common bird species, we Te hazard maps illustrate spatial diferences in opti- directly normalized the abundance data by assigning 100 mal environmental conditions for virus circulation and to cells where the species was most common and 0 where therefore portray relative hazard rather than actual haz- it was absent. ard. Tis relative hazard is expressed on a scale between 0 (low hazard) and 100 (high hazard) and is visualised on the maps with colours ranging from black (low hazard Construction of hazard maps area) to red (high hazard area). Because TBEV emerged We used ‘static risk mapping’ (sensu [82]) to character- in the Netherlands while this study was ongoing, we com- ize the spatial variation in the above-described ecological pared our establishment map with locations where TBEV risk factors for arbovirus circulation. First, we gener- was detected in ticks, wildlife, and humans, and where ated a grid of 5 × 5 km covering the Netherlands in ESRI serologically positive roe deer were found [83]. Tese ArcGIS 10.5 (ESRI 2017). Cells that had their centroids data were obtained from the website of the National > 1 km away from land were excluded from the analy- Institute for Public Health and the Environment (RIVM), ses to prevent edge efects. Each ecological risk factor to which the Dutch Wildlife Health Centre, ErasmusMC, was represented by a single GIS-layer that covered the LabMicTa, MPH Services (GGD), Wageningen Univer- entire grid. Values for each layer were averaged per cell sity and Research, and Artemis One Health contributed. and then normalized (0–100) over the entire grid to allow In addition, we compared our WNV establishment map adding or subtracting in further analyses (see below). with locations where serologically positive birds were All individual layers are provided in Additional fle 3: recently reported [84]. Figures S1–S19. Te introduction maps for CCHFV, WNV and JEV Results consisted of one GIS-layer each, i.e. the relative abun- Te introduction maps highlight locations with large dance of migratory birds (Table 1). In contrast, the estab- concentrations of migratory birds from appropriate lishment maps of the six arboviruses were constructed source areas or species groups, which may either be by combining (overlaying) multiple GIS-layers. For infected with WNV or JEV, or carry CCHFV-infected H. this, we frst classifed each of these layers (risk factors) marginatum ticks from endemic regions (Figs. 1, 2 and as belonging to either abiotic conditions, vector abun- 3). Te establishment maps indicated that for WNV, the dance, or host availability (Table 1). Because the relative southern and western part of the Netherlands are most importance of each risk factor varies between endemic suitable for endemic circulation (Fig. 4), while for JEV regions [22], it remains unclear which factor(s) will con- suitability was highest in the southern and eastern part tribute most to virus circulation in the Netherlands. All of the country (Fig. 5). For RVFV and CCHFV, only few layers (risk factors) within each of the groups (abiotic, locations were classifed as having a relatively high haz- vector and host) were therefore weighted equally in the ard for establishment, but they were all located in the analysis by averaging all values per grid cell. Te three south (Figs. 6, 7). Establishment hazard of TBEV was groups were then overlaid and weighted equally again to highest in nature areas in central, southern, and eastern prevent bias towards one particular group for arbovirus parts of the country, where the combination of seasonal establishment. temperature profles, suitable tick habitat and host availa- For example, in the case of CCHFV, the host layer con- bility are most likely to allow for co-feeding transmission sisted of livestock abundance, the vector layer consisted between infected and uninfected ticks on rodent hosts of suitable habitat for H. marginatum, whereas the abi- (Fig. 8). In contrast, the hazard of LIV establishment was otic layer was a combination of temperature and pre- highest in the north of the country (Fig. 9). Overlaying cipitation (Table 1). As we included a positive efect of all of the establishment maps showed that overall haz- temperature but a negative efect of precipitation, their ard of endemic arbovirus circulation was highest in the values could not simply be averaged. We therefore sub- southern parts of the country; a region characterized by a tracted the precipitation values from 100 to obtain a scale warmer climate, which positively afects vectorial capac- where higher values correspond with lower precipita- ity and vector abundance [56, 75, 85] (Fig. 10). tion (which is favourable for H. marginatum). Tempera- As TBEV recently emerged in the Netherlands, we ture and precipitation values were then averaged per grid compared our establishment maps with the specifc loca- cell and normalized. Te three abiotic, host, and vector tions where the virus has been detected in ticks, humans, Esser et al. Parasites Vectors (2020) 13:464 Page 7 of 20

and wildlife and where seropositive roe deer were found and the frst imported Hyalomma ticks have already been [83]. Locations where PCR-positive ticks, wildlife and found on migratory birds, horses and humans [31, 32]. human cases have been found are all marked as high- Together, these fndings warrant increased surveillance hazard areas, while serological evidence is present in for TBEV, WNV and CCHFV. medium- to high-hazard areas, providing supportive In contrast, the potential introduction of RVFV, JEV evidence for the validity of our maps (Fig. 8). In addi- and LIV is probably more dependent on human activities, tion, WNV-specifc neutralizing antibodies were recently such as trade and travel, rather than natural movement of detected in birds (Eurasian coot Fulica atra and carrion hosts and/or vectors. For example, the red grouse plays crow) from Amsterdam, Rotterdam and Te Hague [84]. an important role in the transmission cycle of LIV in the While it is possible that these birds acquired WNV out- UK [89], but this species is absent in the Netherlands. Its side of the Netherlands, the cities in which they were closest relative, the black grouse Lyrurus tetrix, is criti- found are located in the western part of the country, cally endangered and has only a small population on the where the relative hazard for establishment was highest Sallandse Heuvelrug. Past introductions of LIV into other (Fig. 4). parts of Europe were likely due to international transport of infected sheep [90], and this is also the most plausible Discussion route of introduction for the Netherlands. Subsequent We performed a spatial analysis of ecological risk fac- establishment of LIV is possible through a competent and tors for circulation of six arboviruses (WNV, JEV, RVFV, abundant vector, I. ricinus [68]. Likewise, few bird species CCHFV, TBEV and LIV) to identify areas in the Neth- that arrive in the Netherlands in summer have overlap- erlands with the highest potential for their introduction ping migratory fyways with birds from JEV-endemic and subsequent establishment. We created introduction areas in Asia, resulting in limited potential for introduc- maps for WNV, JEV and CCHFV, and establishment tion when temperatures are suitable for viral replication. maps for each of the six arboviruses (Figs. 1, 2, 3, 4, 5, 6, Such a long-distance migration might also reduce the 7, 8 and 9). We stress that these maps portray spatial vari- viremia of these birds, so that they are no longer infec- ation in relative hazard, i.e. arbovirus circulation is more tious upon arrival [91]. Alternatively, introduction of likely in certain locations than in others, rather than JEV as well as RVFV via infected mosquitoes that come actual hazard. Tat being said, the similarity between with trade or air trafc is theoretically possible [91], but the predicted high-hazard areas and the locations where is considered to be less likely for the Netherlands than actual TBEV-cases and WNV-serologically positive birds entry through (illegal) trade of birds (JEV) and livestock were reported, provides some confdence that our spatial (RVFV) [92]. Access to animal trade data is of critical model, despite its relative simplicity, can be used to iden- importance for mapping this hazard, and it is therefore tify regions in the Netherlands where arbovirus emer- extremely unfortunate that this information was not gence is most likely. made available by the relevant bodies, who deemed it to While TBEV appears to be locally established in the be economically too sensitive. Netherlands, autochthonous WNV infections in mos- Te use of GIS-based models has become increasingly quitoes or animals have yet to be reported, despite common in the feld of spatial epidemiology to map the widespread availability of competent vectors [39] and potential emergence of diseases in areas beyond their expected introduction by viraemic birds [30, 86]. Tem- current distribution [82, 93–96]. Our approach is similar perature is regarded as the key limiting factor for WNV and generated useful results that are corroborated with transmission in northern Europe [56, 87], but the virus is recent fndings of e.g. TBEV emergence. Further, over- predicted to spread into this region via migratory birds lap between some of the arboviruses’ high-hazard areas under future climate change scenarios [88]. Similar pre- supports the implementation of integrated surveillance dictions have been made for H. marginatum, which in regions where multiple arboviruses may emerge. On is both the main vector and reservoir host of CCHFV the other hand, it is difcult to weigh for diferences in in Europe [75]. A population model showed that self- importance or efect sizes of diferent risk factors in our sustaining populations of this tick species were absent analyses at this stage. More experimental research is in areas where yearly accumulated temperatures drop required to disentangle the efects of the diferent, often below 3000–4000 °C [60]. In the Netherlands, the yearly confounded environmental factors and to estimate their accumulated temperature averaged 3751 °C during our relative contribution to virus circulation. However, it study period (2010–2015), but is expected to rise above is relatively easy to include a weighting factor in our 4000 °C under the predicted temperature increase of analyses, or to add additional layers (e.g. risk for human 1 °C by 2030 [13]. However, this temperature limit was exposure via recreation) so that the accuracy of the pre- already exceeded in the exceptionally warm year of 2018, dictions can be improved. Tese predictions also need to Esser et al. Parasites Vectors (2020) 13:464 Page 8 of 20

Fig. 1 Hazard map for the introduction of West Nile virus (WNV) in the Netherlands

be tested, and such a validation phase is a fundamental Testing the accuracy of these predictions will be a chal- requirement to improve our understanding of the under- lenge, as most of the arboviruses considered here are still lying causal mechanisms driving these spatial patterns. presumed absent from the Netherlands (i.e. CCHF, LIV, Esser et al. Parasites Vectors (2020) 13:464 Page 9 of 20

Fig. 2 Hazard map for the introduction of Japanese encephalitis virus (JEV) in the Netherlands

JEV, WNV and RVFV). However, the recent emergence Netherlands shows some resemblance to the establish- of TBEV and the closely to WNV-related Usutu virus ment map for WNV, with a gradual northwest-oriented [15], might ofer some prospect for testing these predic- spread from the southeast of the Netherlands [15]. tions. Indeed, the establishment pattern of USUV in the Esser et al. Parasites Vectors (2020) 13:464 Page 10 of 20

Fig. 3 Hazard map for the introduction of Crimean-Congo haemorrhagic fever virus (CCHFV) in the Netherlands Esser et al. Parasites Vectors (2020) 13:464 Page 11 of 20

Fig. 4 Hazard map for the establishment of West Nile virus (WNV) in the Netherlands. Locations where birds with WNV-neutralizing antibodies were caught [83] are indicated with black circles. Location 1: Amsterdam; Location 2: The Hague; Location 3: Rotterdam Esser et al. Parasites Vectors (2020) 13:464 Page 12 of 20

Fig. 5 Hazard map for the establishment of Japanese encephalitis virus (JEV) in the Netherlands Esser et al. Parasites Vectors (2020) 13:464 Page 13 of 20

Fig. 6 Hazard map for the establishment of Rift Valley fever virus (RVFV) in the Netherlands Esser et al. Parasites Vectors (2020) 13:464 Page 14 of 20

Fig. 7 Hazard map for the establishment of Crimean-Congo haemorrhagic fever virus (CCHFV) in the Netherlands

Conclusions emergence. Our analyses and the generated hazard maps Te use of spatial models has become a key method to show that there is spatial clustering of areas with either a map the environmental suitability for arbovirus circula- relatively low or high potential for arbovirus introduction tion and to target surveillance in regions of potential and/or establishment in the Netherlands. Importantly, Esser et al. Parasites Vectors (2020) 13:464 Page 15 of 20

Fig. 8 Hazard map for the establishment of tick-borne encephalitis virus (TBEV) in the Netherlands. Locations where TBEV-positive ticks, wildlife, and human cases were found are indicated with black circles. Locations where TBEV-seropositive wildlife were found are indicated with dashed circles [83] Esser et al. Parasites Vectors (2020) 13:464 Page 16 of 20

Fig. 9 Hazard map for the establishment of louping-ill virus (LIV) in the Netherlands Esser et al. Parasites Vectors (2020) 13:464 Page 17 of 20

Fig. 10 Combined establishment map for all six arboviruses (WNV, JEV, RVFV, TBEV, CCHFV and LIV) shows that the relative hazard is highest in the southern part of the Netherlands Esser et al. Parasites Vectors (2020) 13:464 Page 18 of 20

some of these high-hazard areas overlap. Our combined upon reasonable request and with permission of the parties from which these data were obtained. map, showing the summed hazard for all six arboviruses per cell, shows that overall hazard is highest in the south- Ethics approval and consent to participate ern part of the country. Sampling of vectors and sentinel Not applicable. hosts should be focused in these key priority areas, where Consent for publication several arboviruses may emerge. Such targeted sampling Not applicable. increases the efcient use of limited resources for surveil- Competing interests lance. Tus, the construction and subsequent overlaying The authors declare that they have no competing interests. of multiple hazard maps provides a promising approach for an integrated, cost-efcient, multiplex-surveillance Author details 1 Wildlife Ecology & Conservation Group, Wageningen University & Research, strategy that targets multiple arboviruses simultaneously. Wageningen, The Netherlands. 2 Laboratory of Entomology, Wageningen University & Research, Wageningen, The Netherlands. 3 Centre for Infectious Supplementary information Disease Control, National Institute for Public Health and the Environment, Bilthoven, The Netherlands. 4 Centre for Monitoring of Vectors (CMV), National Supplementary information accompanies this paper at https​://doi. Reference Centre (NRC), Netherlands Food and Consumer Product Safety org/10.1186/s1307​1-020-04339​-0. Authority (NVWA), Ministry of Agriculture, Nature and Food Quality, Wagen‑ ingen, The Netherlands. 5 Vogeltrekstation - Dutch Centre for Avian Migration 6 Additional fle 1: Table S1. Bird species that migrate from Africa and/or and Demography (NIOO-KNAW), Wageningen, The Netherlands. Sovon 7 the Mediterranean area to North-western Europe during spring. Table S2. Dutch Centre for Field Ornithology, Nijmegen, The Netherlands. Department Bird species that have overlapping migratory fyways with conspecifcs of Animal Ecology & Ecophysiology, Institute for Water and Wetland Research, 8 migrating from JEV-endemic regions in southeast Asia. Table S3. Ardeid Radboud University, Nijmegen, The Netherlands. Department of Viroscience, bird species of the Netherlands. Table S4. Wetland bird species of the WHO CC for Arbovirus and Viral Hemorrhagic Fever Reference and Research, Netherlands. Table S5. Land cover classes of the Netherlands (LGN7) that Erasmus University Medical Centre, Rotterdam, The Netherlands. were included as habitat for Ixodes ricinus and Hyalomma marginatum ticks. Habitat suitability ranged from 0 (not suitable) to 3 (very suitable). Received: 13 April 2020 Accepted: 1 September 2020 Additional fle 2: Text S1. Description of the methods used for modelling mosquito abundance, including information on mosquito data collection, the environmental data used and the statistical methods applied. Additional fle 3: Figures S1–S19. Individual layers (ecological risk fac‑ References tors) that were used for constructing the hazard maps. 1. Sigfrid L, Reusken C, Eckerle I, Nussenblatt V, Lipworth S, Messina J, et al. Preparing clinicians for (re-)emerging arbovirus infectious diseases in Additional fle 4: Figures S20–S25. Schematic representation of the Europe. Clin Microbiol Infect. 2017;24:229–39. steps taken to construct the hazard maps for the establishment of each of 2. Süss J. Tick-borne encephalitis 2010: epidemiology, risk areas, and the six arboviruses. virus strains in Europe and Asia—an overview. Ticks Tick Borne Dis. 2011;2:2–15. Abbreviations 3. Papa A. Emerging arboviral human diseases in southern Europe. J Med CCHFV: Crimean-Congo haemorrhagic fever virus; JEV: Japanese encepha‑ Virol. 2017;89:1315–22. litis virus; LIV: Louping-ill virus; RVFV: Rift Valley fever virus; TBEV: Tick-borne 4. Holding M, Dowall S, Medlock J, Carter D, Pullan S, Lewis J, et al. Tick- encephalitis virus; WNV: West Nile virus; KNMI: Royal Netherlands Meteorologi‑ borne encephalitis virus, United Kingdom. Emerg Infect Dis. 2020;26:90–6. cal Institute; RH: 24-h sum of precipitation; TG: 24-h average temperature; TX: 5. de Wispelaere M, Desprès P, Choumet V. European Aedes albopictus and 24-h maximum temperature; UG: 24-h average relative humidity. Culex pipiens are competent vectors for Japanese encephalitis virus. PLoS Negl Trop Dis. 2017;11:e0005294. Acknowledgements 6. Rolin AI, Berrang-Ford L, Kulkarni MA. The risk of Rift Valley fever virus We thank the following organisations and institutes for providing raw source introduction and establishment in the United States and European data: Royal Netherlands Meteorological Institute (KNMI), Netherlands Enter‑ Union. Emerg Microbes Infect. 2013;2:e81. prise Agency (RVO), Wageningen Environmental Research, Dutch National 7. Šumilo D, Bormane A, Asokliene L, Vasilenko V, Golovljova I, Avsic-Zupanc Database Flora and Fauna; and the Dutch Centre for Field Ornithology T, et al. Socio-economic factors in the diferential upsurge of tick-borne (SOVON). encephalitis in central and eastern Europe. Rev Med Virol. 2008;18:81–95. 8. Estrada-Peña A, Vatansever Z, Gargili A, Ergönul Ö. The trend towards Authors’ contributions habitat fragmentation is the key factor driving the spread of Crimean- WFB, CBEMR and MPGK conceived the study. HJE identifed the ecological risk Congo haemorrhagic fever. Epidemiol Infect. 2010;138:1194–203. factors and YL obtained the associated raw source data. AIJ, HJ, CAMT and AS 9. Webster JL, Stapleford KA. Arbovirus adaptation: roles in transmission and provided input data. YL performed the analyses and created the hazard maps. emergence. Curr Clin Microbiol Rep. 2017;4:159–66. HJE, CBEMR and WFB wrote the manuscript with input from all authors. All 10. Vogels CB, Hartemink N, Koenraadt CJ. Modelling West Nile virus trans‑ authors read and approved the fnal manuscript. mission risk in Europe: efect of temperature and mosquito biotypes on the basic reproduction number. Sci Rep. 2017;7:5022. Funding 11. Havelaar AH, Van Rosse F, Bucura C, Toetenel MA, Haagsma JA, Kurowicka This study was fnancially support by a ZonMw grant (522001004). D, et al. Prioritizing emerging zoonoses in the Netherlands. PLoS ONE. 2010;5:e13965. Availability of data and materials 12. Gould EA, Higgs S. Impact of climate change and other factors on emerg‑ The abiotic data that support the fndings of this study are available from the ing arbovirus diseases. Trans R Soc Trop Med Hyg. 2009;103:109–21. Royal Netherlands Meteorological Institute (KNMI). However, restrictions apply 13. van den Hurk B, Siegmund P, Klein Tank A. KNMI’14: climate change sce‑ to the availability of the host and vector data (e.g. abundance of livestock and narios for the 21st century—a Netherlands perspective. Scientifc Report birds, land-use dataset), which were used under license for the current study, WR2014-01. De Bilt, the Netherlands: KNMI; 2014. www.clima​tesce​nario​ and so are not publicly available. Data are however available from the authors s.nl. Esser et al. Parasites Vectors (2020) 13:464 Page 19 of 20

14. Dekker M, Laverman GD, de Vries A, Reimerink J, Geeraedts F. Emergence 37. Kazarina A, Japiņa K, Keišs O, Salmane I, Bandere D, Capligina V, et al. of tick-borne encephalitis (TBE) in the Netherlands. Ticks Tick Borne Dis. Detection of tick-borne encephalitis virus in I. ricinus ticks col‑ 2019;10:176–9. lected from autumn migratory birds in Latvia. Ticks Tick Borne Dis. 15. Rijks JM, Kik ML, Slaterus R, Foppen RPB, Stroo A, IJzer J, et al. Widespread 2015;6:178–80. Usutu virus outbreak in birds in the Netherlands 2016. Euro Surveill. 38. Braks MAH, Stroo CJ. The control of native mosquitoes in the Nether‑ 2016;21:30391. lands: possibilities and challenges. RIVM report 2016-0004. The Bilt, 16. Zannoli S, Morotti M, Denicolò A, Tassinari M, Chiesa C, Pierro A, et al. Chi‑ the Netherlands: RIVM; 2016. https​://www.rivm.nl/bibli​othee​k/rappo​ kungunya virus and Zika virus in Europe. In: Higgs S, Vanlandingham DL, rten/2016-0004.pdf. Powers A, editors. Chikungunya and zika viruses: global emerging health 39. Fros JJ, Geertsema C, Vogels CB, Roosjen PP, Failloux AB, Vlak JM, et al. threats. London: Elsevier; 2018. p. 193–214. West Nile virus: high transmission rate in north-western European 17. Cadar D, Lühken R, van der Jeugd H, Garigliany M, Ziegler U, Keller M, mosquitoes indicates its epidemic potential and warrants increased et al. Widespread activity of multiple lineages of Usutu virus, western surveillance. PLoS Negl Trop Dis. 2015;9:e0003956. Europe, 2016. Euro Surveill. 2017;22:30452. 40. Vloet RPM, Vogels CBF, Koenraadt CJM, Pijlman GP, Eiden M, Gon‑ 18. Nash D, Mostashari F, Fine A, Miller J, O’leary D, Murray K, et al. The zales JL, et al. Transmission of Rift Valley fever virus from European- outbreak of West Nile virus infection in the New York City area in 1999. N breed lambs to Culex pipiens mosquitoes. PLoS Negl Trop Dis. Engl J Med. 2001;344:1807–14. 2017;11:e0006145. 19. ECDC. Aedes albopictus—current known distribution in Europe, April 41. Ibañez-Justicia A, Cianci D. Modelling the spatial distribution of the 2017. Stockholm, Sweden: European Centre for Disease Prevention and nuisance mosquito species Anopheles plumbeus (Diptera: Culicidae) in the Control; 2018. https​://ecdc.europ​a.eu/en/disea​se-vecto​rs/surve​illan​ce- Netherlands. Parasites Vectors. 2015;8:258. and-disea​se-data/mosqu​ito-maps. 42. Ibañez-Justicia A, Stroo A, Dik M, Beeuwkes J, Scholte EJ. National mos‑ 20. Kalluri S, Gilruth P, Rogers D, Szczur M. Surveillance of arthropod vector- quito (Diptera: Culicidae) survey in The Netherlands 2010–2013. J Med borne infectious diseases using remote sensing techniques: a review. Entomol. 2015;52:185–98. PLoS Pathog. 2007;3:e116. 43. Bird BH, Ksiazek TG, Nichol ST, MacLachlan NJ. Rift Valley fever virus. J Am 21. Eisen RJ, Eisen L. Spatial modeling of human risk of exposure to vector- Vet Med Assoc. 2009;234:883–93. borne pathogens based on epidemiological versus arthropod vector 44. Lubisi BA, Ndouvhada PN, Neifer D, Penrith ML, Sibanda DR, Bastos A. data. J Med Entomol. 2008;45:181–92. Seroprevalence of Rift valley fever in South African domestic and wild 22. Esser HJ, Mögling R, Cleton NB, van der Jeugd H, Sprong H, Stroo A, suids (1999–2016). Transbound Emerg Dis. 2020;67:811–21. et al. Risk factors associated with sustained circulation of six zoonotic 45. Bird BH, McElroy AK. Rift Valley fever virus: unanswered questions. Antivir arboviruses: a systematic review for selection of surveillance sites in non- Res. 2016;132:274–80. endemic areas. Parasites Vectors. 2019;12:265. 46. Rostal MK, Liang JE, Zimmermann D, Bengis R, Paweska J, Karesh WB. Rift 23. Messina JP, Kraemer MU, Brady OJ, Pigott DM, Shearer FM, Weiss DJ, Valley fever: does wildlife play a role? ILAR J. 2017;58:359–70. et al. Mapping global environmental suitability for Zika virus. Elife. 47. Rizzoli A, Bolzoni L, Chadwick EA, Capelli G, Montarsi F, Grisenti M, et al. 2016;5:e15272. Understanding West Nile virus ecology in Europe: Culex pipiens host 24. Nsoesie EO, Kraemer MU, Golding N, Pigott DM, Brady OJ, Moyes CL, et al. feeding preference in a hotspot of virus emergence. Parasites Vectors. Global distribution and environmental suitability for chikungunya virus, 2015;8:213. 1952 to 2015. Euro Surveill. 2016. https​://doi.org/10.2807/1560-7917. 48. Hubálek Z, Halouzka J. West Nile fever—a reemerging mosquito-borne es.2016.21.20.30234​. viral disease in Europe. Emerg Infect Dis. 1999;5:643. 25. Racloz V, Presi P, Vounatsou P, Schwermer H, Casati S, Vanzetti T, et al. 49. Rappole JH, Derrickson SR, Hubálek Z. Migratory birds and spread of West Use of mapping and statistical modelling for the prediction of blue‑ Nile virus in the western Hemisphere. Emerg Infect Dis. 2000;6:319. tongue occurrence in Switzerland based on vector biology. Vet Ital. 50. Lim SM, Brault AC, van Amerongen G, Bosco-Lauth AM, Romo H, 2007;43:513–8. Sewbalaksing VD, et al. Susceptibility of carrion crows to experimen‑ 26. Messina JP, Pigott DM, Golding N, Duda KA, Brownstein JS, Weiss DJ, et al. tal infection with lineage 1 and 2 West Nile viruses. Emerg Infect Dis. The global distribution of Crimean-Congo hemorrhagic fever. Trans R Soc 2015;21:1357. Trop Med Hyg. 2015;109:503–13. 51. Del Amo J, Llorente F, Figuerola J, Soriguer RC, Moreno AM, Cordioli P, 27. Anyamba A, Chretien JP, Small J, Tucker CJ, Formenty PB, Richardson JH, et al. Experimental infection of house sparrows (Passer domesticus) with et al. Prediction of a Rift Valley fever outbreak. Proc Natl Acad Sci USA. West Nile virus isolates of Euro-Mediterranean and North American 2009;106:955–9. origins. Vet Res. 2014;45:33. 28. Pfefer M, Dobler G. Emergence of zoonotic arboviruses by animal trade 52. Lim SM, Brault AC, van Amerongen G, Sewbalaksing VD, Osterhaus AD, and migration. Parasites Vectors. 2010;3:35. Martina BE, et al. Susceptibility of European jackdaws (Corvus monedula) 29. Jameson LJ, Morgan PJ, Medlock JM, Watola G, Vaux AGC. Importation of to experimental infection with lineage 1 and 2 West Nile viruses. J Gen Hyalomma marginatum, vector of Crimean-Congo haemorrhagic fever Virol. 2014;95:1320–9. virus, into the United Kingdom by migratory birds. Ticks Tick Borne Dis. 53. Jiménez de Oya N, Camacho MC, Blázquez AB, Lima-Barbero JF, Saiz JC, 2012;3:95–9. Höfe U, et al. High susceptibility of magpie (Pica pica) to experimen‑ 30. Rappole JH, Hubalek Z. Migratory birds and West Nile virus. J Appl Micro‑ tal infection with lineage 1 and 2 West Nile virus. PLoS Negl Trop Dis. biol. 2003;94:47–58. 2018;12:e0006394. 31. RIVM. Zoonotic diseases report 2012. The Bilt, the Netherlands: National 54. Spedicato M, Carmine I, Bellacicco AL, Marruchella G, Marini V, Pisciella M, Institute for Public Health and the Environment; 2013. https​://www.rivm. et al. Experimental infection of rock pigeons (Columba livia) with three nl/bibli​othee​k/rappo​rten/09233​0002.pdf. West Nile virus lineage 1 strains isolated in Italy between 2009 and 2012. 32. Butler CM, van Oldruitenborgh-Oosterbaan MMS, Stout TA, Jongejan F, Epidemiol Infect. 2016;144:1301–11. Werners AH, Houwers DJ. Classifcation of ticks collected from horses in 55. Kilpatrick AM, Meola MA, Moudy RM, Kramer LD. Temperature, viral the Netherlands in 2008–2009 and identifcation of the (zoonotic) agents genetics, and the transmission of West Nile virus by Culex pipiens mosqui‑ they contain. Pferdeheilkunde. 2016;32:329. toes. PLoS Pathog. 2008;4:e1000092. 33. Misra UK, Kalita J. Overview: Japanese encephalitis. Prog Neurobiol. 56. Vogels CBF, Göertz GP, Pijlman GP, Koenraadt CJM. Vector competence 2010;91:108–20. of northern and southern European Culex pipiens pipiens mosquitoes 34. Chevalier V, Pépin M, Plée L, Lancelot R. Rift Valley fever—a threat for for West Nile virus across a gradient of temperatures. Med Vet Entomol. Europe? Euro Surveill. 2010;15:19506. 2017;34:358–64. 35. Ergonul O, Whitehouse CA. Crimean-Congo hemorrhagic fever: a global 57. Bi P, Zhang Y, Parton KA. Weather variables and Japanese encephalitis in perspective. Dordrecht: Springer Science & Business Media; 2007. the metropolitan area of Jinan city, China. J Infect. 2007;55:551–6. 36. Jaenson TG, Hjertqvist M, Bergström T, Lundkvist Å. Why is tick-borne 58. Marcantonio M, Rizzoli A, Metz M, Rosà R, Marini G, Chadwick E, et al. encephalitis increasing? A review of the key factors causing the increas‑ Identifying the environmental conditions favouring West Nile virus ing incidence of human TBE in Sweden. Parasites Vectors. 2012;5:184. outbreaks in Europe. PLoS ONE. 2015;10:e0121158. Esser et al. Parasites Vectors (2020) 13:464 Page 20 of 20

59. Turell MJ, Rossi CA, Bailey CL. Efect of extrinsic incubation temperature 81. SOVON, Dutch centre for feld ornithology. Bird atlas Netherlands: breed‑ on the ability of Aedes taeniorhynchus and Culex pipiens to transmit Rift ing and wintering birds and 40 years of change. Utrecht/Antwerpen: Valley fever virus. Am J Trop Med Hyg. 1985;34:1211–8. Kosmos Uitgevers; 2018. 60. Estrada-Peña A, Martínez Avilés M, Muñoz Reoyo MJ. A population 82. Ostfeld RS, Glass GE, Keesing F. Spatial epidemiology: an emerging (or model to describe the distribution and seasonal dynamics of the tick re-emerging) discipline. Trends Ecol Evol. 2005;20:328–36. Hyalomma marginatum in the Mediterranean Basin. Transbound Emerg 83. RIVM. Spread of tick-borne encephalitis virus in the Netherlands. The Bilt, Dis. 2011;58:213–23. the Netherlands: RIVM; 2020. https​://www.rivm.nl/en/news/sprea​d-of- 61. Lindström A, Jaenson TGT. Distribution of the common tick, Ixodes ricinus tick-borne​-encep​halit​is-virus​-in-nethe​rland​s. Accessed 6 July 2020. (Acari: Ixodidae), in diferent vegetation types in southern Sweden. J Med 84. Lim SM, Geervliet M, Verhagen JH, Müskens G, Majoor FA, Osterhaus Entomol. 2003;40:375–8. AD, et al. Serologic evidence of West Nile virus and Usutu virus infec‑ 62. Uspensky I. Preliminary observations on specifc adaptations of exophilic tions in Eurasian coots in the Netherlands. Zoonoses Public Health. ixodid ticks to forests or open country habitats. Exp Appl Acarol. 2018;65:96–102. 2002;28:147–54. 85. Miller RH, Masuoka P, Klein TA, Kim H-C, Somer T, Grieco J. Ecological 63. Rizzoli A, Haufe HC, Tagliapietra V, Neteler M, Rosà R. Forest structure niche modeling to estimate the distribution of Japanese encephalitis and roe deer abundance predict tick-borne encephalitis risk in Italy. PLoS virus in Asia. PLoS Negl Trop Dis. 2012;6:e1678. ONE. 2009;4:e4336. 86. Ziegler U, Lühken R, Keller M, Cadar D, van der Grinten E, Michel F, et al. 64. Kifner C, Zucchini W, Schomaker P, Vor T, Hagedorn P, Niedrig M, et al. West Nile virus epizootic in Germany, 2018. Antivir Res. 2019;162:39–43. Determinants of tick-borne encephalitis in counties of southern Ger‑ 87. Paz S. Climate change impacts on West Nile virus transmission in a global many, 2001–2008. Int J Health Geogr. 2010;9:42. context. Philos Trans R Soc B. 2015;370:20130561. 65. Zeimes CB, Olsson GE, Hjertqvist M, Vanwambeke SO. Shaping zoonosis 88. Semenza JC, Tran A, Espinosa L, Sudre B, Domanovic D, Paz S. Climate risk: landscape ecology vs. landscape attractiveness for people, the case change projections of West Nile virus infections in Europe: implications of tick-borne encephalitis in Sweden. Parasites Vectors. 2014;7:370. for blood safety practices. Environ Health. 2016;15:S28. 66. Vanwambeke SO, Šumilo D, Bormane A, Lambin EF, Randolph SE. Land‑ 89. Laurenson MK, Norman RA, Gilbert L, Reid HW, Hudson PJ. Identifying scape predictors of tick-borne encephalitis in Latvia: land cover, land use, disease reservoirs in complex systems: mountain hares as reservoirs and land ownership. Vector-Borne Zoonotic Dis. 2010;10:497–506. of ticks and louping-ill virus, pathogens of red grouse. J Anim Ecol. 67. Vescio FM, Busani L, Mughini-Gras L, Khoury C, Avellis L, Taseva E, et al. 2003;72:177–85. Environmental correlates of Crimean-Congo haemorrhagic fever inci‑ 90. McGuire K, Holmes EC, Gao GF, Reid HW, Gould EA. Tracing the origins dence in Bulgaria. BMC Public Health. 2012;12:1. of louping ill virus by molecular phylogenetic analysis. J Gen Virol. 68. Gilbert L. Louping ill virus in the UK: a review of the hosts, trans‑ 1998;79:981–8. mission and ecological consequences of control. Exp Appl Acarol. 91. van den Hurk AF, Ritchie SA, Mackenzie JS. Ecology and geographi‑ 2016;68:363–74. cal expansion of Japanese encephalitis virus. Annu Rev Entomol. 69. Hofmeester TR, Sprong H, Jansen PA, Prins HHT, van Wieren SE. Deer pres‑ 2009;54:17–35. ence rather than abundance determines the population density of the 92. EFSA Panel on Animal Health and Welfare (AHAW). Scientifc opinion on sheep tick, Ixodes ricinus, in Dutch forests. Parasites Vectors. 2017;10:433. rift valley fever. EFSA J. 2013;11:3180. 70. van Wieren SE, Hofmeester TR. The role of large herbivores in Ixodes 93. Ogden NH, St-Onge L, Barker IK, Brazeau S, Bigras-Poulin M, Charron DF, ricinus and Borrelia burgdorferi sl dynamics. In: Braks MAH, van Wieren SE, et al. Risk maps for range expansion of the Lyme disease vector, Ixodes Takken W, Sprong H, editors. Ecology and prevention of Lyme borreliosis scapularis, in Canada now and with climate change. Int J Health Geogr. and vector-borne diseases. Wageningen: Wageningen Academic Publish‑ 2008;7:24. ers; 2016. p. 75–89. 94. Guerra M, Walker E, Jones C, Paskewitz S, Cortinas MR, Stancil A, et al. 71. Randolph SE. Transmission of tick-borne pathogens between co- Predicting the risk of Lyme disease: habitat suitability for Ixodes scapularis feeding ticks: Milan Labuda’s enduring paradigm. Ticks Tick Borne Dis. in the north central United States. Emerg Infect Dis. 2002;8:289–97. 2011;2:179–82. 95. Brooker S. Spatial epidemiology of human schistosomiasis in Africa: risk 72. Jones LD, Gaunt M, Hails RS, Laurenson K, Hudson PJ, Reid H, et al. models, transmission dynamics and control. Trans R Soc Trop Med Hyg. Transmission of louping ill virus between infected and uninfected ticks 2007;101:1–8. co-feeding on mountain hares. Med Vet Entomol. 1997;11:172–6. 96. de la Torre A, Bosch J, Iglesias I, Muñoz MJ, Mur L, Martínez-López B, et al. 73. Medlock JM, Hansford KM, Bormane A, Derdakova M, Estrada-Peña A, Assessing the risk of African swine fever introduction into the European George JC, et al. Driving forces for changes in geographical distribution Union by wild boar. Transbound Emerg Dis. 2015;62:272–9. of Ixodes ricinus ticks in Europe. Parasites Vectors. 2013;6:1. 74. Estrada-Peña A, Bouattour A, Camicas JL, Walker AR. Ticks of domestic animals in the Mediterranean region: a guide to identifcation of species. Publisher’s Note Zaragoza: University of Zaragoza; 2004. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 75. Estrada-Peña A, Sánchez N, Estrada-Sánchez A. An assessment of the lished maps and institutional afliations. distribution and spread of the tick Hyalomma marginatum in the western Palearctic under diferent climate scenarios. Vector Borne Zoonotic Dis. 2012;12:758–68. 76. Estrada-Peña A, Venzal JM. Climate niches of tick species in the Mediter‑ ranean region: modeling of occurrence data, distributional constraints, and impact of climate change. J Med Entomol. 2007;44:1130–8. Ready to submit your research ? Choose BMC and benefit from: 77. Williams HW, Cross DE, Crump HL, Drost CJ, Thomas CJ. Climate suitability • fast, convenient online submission for European ticks: assessing species distribution models against null models and projection under AR5 climate. Parasites Vectors. 2015;8:440. • thorough peer review by experienced researchers in your field 78. Sluiter R. Interpolation methods for the climate atlas. Technical report; • rapid publication on acceptance TR-335. The Bilt, the Netherlands: KNMI; 2012. http://bibli​othee​k.knmi.nl/ • support for research data, including large and complex data types knmip​ubTR/TR335​.pdf. 79. Hazeu GW, Schuiling C, Dorland GJ, Roerink GJ, Naef HSD, Smidt RA. • gold Open Access which fosters wider collaboration and increased citations Landelijk Grondgebruiksbestand Nederland versie 7 (LGN7): manufacture, • maximum visibility for your research: over 100M website views per year accuracy, and use. Wageningen: Alterra Wageningen UR; 2014. http:// conte​nt.alter​ra.wur.nl/Webdo​cs/PDFFi​les/Alter​rarap​porte​n/Alter​raRap​ At BMC, research is always in progress. port2​548.pdf. 80. NDFF. NDFF Verspreidingsatlas. 2018. https​://www.versp​reidi​ngsat​las.nl/. Learn more biomedcentral.com/submissions